GAT-YOLOv11-Based Classification Method for Pediatric Wrist Fractures

Precise diagnosis and meticulous classification are paramount to the effective treatment of pediatric distal radius fractures. Traditional X-ray exams are often difficult to interpret and have a high misdiagnosis rate. This research introduces a novel approach that employs an enhanced YOLOv11 algorithm and the GRAZPEDWRI-DX dataset, incorporating data augmentation techniques. The model includes CGD and DAT modules to enhance context awareness and focus on important areas. The improved model achieves an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{m A P}$</tex> of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{6 6. 2 \%}$</tex> at <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{I o U}=\mathbf{0. 5}$</tex> for pediatric wrist fracture classification, outperforming comparison models. Ablation studies confirm the effectiveness of the modules. This approach significantly improves diagnostic accuracy and efficiency, supporting clinical treatment.